aws doubles down on custom llms with features meant to simplify model creation

AWS Expands Frontier Model Tools for Enterprises

Following the unveiling of Nova Forge, a service dedicated to the training of bespoke Nova AI models, Amazon Web Services (AWS) has revealed further resources designed for enterprise clients seeking to develop their own cutting-edge models.

New Capabilities in Bedrock and SageMaker

The announcements, made at the AWS re:Invent conference on Wednesday, introduce enhanced functionalities within Amazon Bedrock and Amazon SageMaker AI. These advancements aim to simplify the process of constructing and refining custom large language models (LLMs) for developers.

According to Ankur Mehrotra, general manager of AI platforms at AWS, the cloud provider is introducing serverless model customization in SageMaker. This allows developers to initiate model building without the need to manage compute resources or underlying infrastructure.

Serverless Model Building Options

Developers can access these serverless capabilities through two distinct pathways: a self-guided, point-and-click interface, or an agent-led experience utilizing natural language prompts to direct SageMaker. The agent-led feature is currently available in preview.

“For instance, a healthcare provider aiming to improve a model’s understanding of specialized medical terminology can simply direct SageMaker AI, providing labeled data, selecting the appropriate technique, and initiating the model fine-tuning process,” Mehrotra explained.

This functionality extends to customizing Amazon’s Nova models, as well as select open source models with publicly accessible model weights, such as DeepSeek and Meta’s Llama.

Reinforcement Fine-Tuning in Bedrock

AWS is also introducing Reinforcement Fine-Tuning within Bedrock. This feature empowers developers to select either a reward function or a pre-defined workflow, allowing Bedrock to autonomously manage the entire model customization process.

Focus on Frontier LLMs

The development of frontier LLMs – representing the most sophisticated AI models – and the ability to customize them appear to be central to AWS’s strategy at this year’s conference.

The introduction of Nova Forge, a service offering custom Nova model development for enterprises at a cost of $100,000 annually, was announced during AWS CEO Matt Garman’s keynote address on Tuesday.

Addressing Customer Differentiation

“Many of our customers are asking how they can distinguish themselves when competitors have access to the same models,” Mehrotra stated. “They want to build unique solutions optimized for their brand, data, and specific use cases.” He emphasized that customized models are key to achieving this differentiation.

Competitive Landscape and Future Prospects

Currently, AWS has not yet established a significant user base for its AI models. A July survey conducted by Menlo Ventures indicated a strong preference among enterprises for Anthropic, OpenAI, and Gemini. However, the enhanced ability to customize and fine-tune LLMs could potentially provide AWS with a competitive edge.

Comprehensive coverage of TechCrunch’s reporting on the annual enterprise tech event can be found here, and a summary of announcements can be accessed here.

Sponsored: Experience AWS re:Invent 2025 Live

Stay informed with the newest announcements concerning topics like agentic AI and cloud infrastructure. Discover advancements in security and a wide range of other innovations.

These revelations are being presented at the premier Amazon Web Services event, taking place in Las Vegas.

This video content is presented in collaboration with AWS.

re:Invent 2025 offers a comprehensive look at the future of cloud computing.

Attendees and viewers alike can gain valuable insights into cutting-edge technologies.

Key Areas of Focus

• Agentic AI: Explore the evolving landscape of artificial intelligence agents.
• Cloud Infrastructure: Learn about the latest developments in scalable and reliable cloud solutions.
• Security: Understand the newest strategies for protecting data and systems.
• Emerging Technologies: Discover innovations shaping the future of the cloud.

The event provides a platform for developers, architects, and IT professionals to connect and learn.

It’s an opportunity to explore how AWS services can drive business transformation.